PDA

View Full Version : anyone who knows much about electronics...

tango
06-01-2003, 11:19 AM
I have a big blue LED, and I want to be able to hook it up to the HDD indicator light, so it flashes and lights up the inside of my box and stuff (I have a window). Anyway I tried connecting the LED straight to the light output on the motherboard, but the voltage coming out is just under 4 volts. The LED's typical operating voltage is 10.5 volts, so I was thinking, I could get a transistor, and use it as a switch, and run the LED from one of the outputs of my power supply. Does anyone know what kind of transistor to use? and how to get it going? and also, is it ok to run the power supply connector, and the HDD light connector onto the same ground line? thanks.

~~~~~ s y ~~~~~
06-01-2003, 11:25 AM
Is there such profession as a Computer Electrician? Maybe, you could try all of those people! ;-)

godfather
06-01-2003, 11:40 AM
LED's have to have the current limited by a resistor, otherwise the LED will instantly self destruct. You wont see any light, even briefly

I am surprised by your comment of 10.5 volts, as most LEDs are in 3 - 4 volt range. I wonder if yours already has a series resistor internally??.

YOU REALLY NEED TO KNOW!

Most LED's require forward current of around 20 mA (milliamps, 0.02 Amps) for nominal brightness, any more can also cause failure. The voltage required is therefore dictated by the value of the resistor and the forward current.

20 mA from 12v supply, (assuming a 3.6v forward drop in the LED), means your resistor should allow 20 mA to flow with (12-3.6) = 8.4 v

R=E(volts)/I(current)

R (resistor value) = 8.4 (volts)/ .02 (20 mA) = 420 Ohms.
Erring on the safe side, the nearest standard value is 470 ohms.

However if the LED already has an internal resistor (its just that I have never seen one that has, but its possible) you need to know what its value is.

Remember NEVER CONNECT AN LED DIRECTLY TO THE SUPPLY, WITHOUT A CURRENT LIMITING DEVICE IN THE CIRCUIT.
Unless you want to buy another LED

Chilling_Silence
06-01-2003, 11:53 AM
*Listen to that 'Whooosh' sound as all that flies right over Chilling_Silence's head!*

I would say that a New LED would be in order, as most are about 3V, but there are other things to watch out for as godfather has probably described in his post :-)

tango
06-01-2003, 12:07 PM
haha yeah I guess I should have described the LED a little more...

http://www.dse.co.nz/cgi-bin/dse.storefront/3e18ba8c0213ea6c2742c0a87f990716/Product/View/Z4230

that's it there. if you were to cut a mouse ball straight down the middle, one half of it would be about the size of the LED. I don't see anything about an internal resistor, but I was going to drop the voltage down to maybe 11 instead of 12 anyway.

Chilling_Silence
06-01-2003, 12:13 PM
Err.... URL seems to be a dud, taking you to the main page as per normal for DSE...

tango
06-01-2003, 12:18 PM

then click the photo to get a full description.

godfather
06-01-2003, 12:22 PM
OK.
It may or may not have an internal resistor, but its the CURRENT you MUST not exceed. If indeed it has just a high forward voltage, and no resistor, connecting 11v will still instantly destroy it. So will 10.5 probably.

Now, as for lighting the case - well, it has a luminance of 30 Mcd.
High brightness LEDs now have a luminance of 20,000 Mcd.
Dont expect it to be very bright.
I doubt it will show any "illumination of the case", you will be able to tell if its on or off though, by looking at the LED itself.

Play safe, use a resistor for a start, try 150 ohms, 1/4 watt rating.

tango
06-01-2003, 12:33 PM
no no..I've already had it going inside my case. I've got it temporily connected to a 9 volt adapter, and it lights up the inside nicely. but I'll check out that stuff about current...

godfather
06-01-2003, 12:40 PM
Look at some of the superbright ones, they are only 5mm in size, but can be used as illumination in torches.

They will certainly light your case, even in daylight, but must have a resistor in series.

I do suspect your blue one has a resistor internally. Otherwise it would be dead. Its not common to have the resistor internal though so be warned if you try another type.

I have torches here that use superbright LEDs, (6,000 Mcd) they are great, batteries last a long time. They require special electronics in the torch though.

tango
06-01-2003, 12:58 PM
ok cool. do you know how to go about getting it connected to the HDD light output?

godfather
06-01-2003, 01:48 PM
Not without measuring it with a multimeter, as it will be one of two ways.

1. The "ground" connection will be switched by the HDD electronics

2. The "supply" connection will be switched by the HDD electronics.

You need a multimeter to determine which.

If you use a cheap small signal NPN transistor, connected with the emitter to ground, and the LED -ve connected to the collector. Connect the LED +ve to a 150 ohm resistor, and the other end of the resistor to +12 v.

Connect a 2,200 ohm resistor to the base lead of the transistor. The other end of the resistor to the HDD LED.

Depending how its driven, one side of the HDD LED will always give blue LED "ON" or "OFF"

The other side of the HDD LED may give blue LED "ON" when HDD LED is on, or "OFF" when HDD LED is on. If its in reverse you will have to get a bit more creative and invert the signal, but I don't have any data to tell me which way it is.
You want the 2,200 ohm resistor to be connected to a signal at the LED on the HDD that goes high (+12v) when the HDD LED is on for the above to work.

All risks are yours, not just to the LED but also to the HDD itself.

Graham L
06-01-2003, 03:21 PM
Been there, done that. Yes it does have a Vf of 10.5 . So assuming you use a transistor "switch" and run it off 12V, a 47 ohm resistor will allow it to pull its 25mA.

It must have a bunch of junctions in series to get a Vf like that --- and the 140 degree visibility ... normally 20 degrees, but that probably explains the 30 mCd. The high intensity ones gain quite a bit bu having a very narrow beam.

So a 10k, a 47R, a BC547 should do you. Cathode of the LED to +12 through 47 ohms, anode to collector of transistor, emitter to ground. 10k resistor between the base of transistor and the disk activity LED. (Usually these supply 5 V through a current limit resistor to the front panel LED.)

My way of driving LEDs from "random" supplies is by putting one of the flashing ones in series. The flashing LEDs have an IC which takes up to 15 V, and incorporates a constant current supply, so you can have a string in series.

andy
06-01-2003, 06:47 PM
Tango,
Spend \$2.50 on a DSE catalogue if you haven't already got one and turn to page 316. Read the section on opto couplers. If you already have the big LED working, then hook up the input to an opto-coupler to the HD led connection - you can either replace it or hook it in series. Connect your new LED circuit to the output side of the optocoupler. Connect the circuit that you have working n series with the collector-emitter circuit of the optocoupler (make sure you get the polarity right). As others have mentioned, you may need some current limiting in the form of a series resistor if you are running your 10.5 Volt led off the 12 volt rail although there is a 1.5 v forward volt drop across the coupler so it should just work. The 4n25 or 4n28 should do the job okay. Also check out "using LEDs" on page 311.
For \$2, the DSE catalogue is a really good, simple electronics textbook.
Regards,
Andy

tango
06-01-2003, 07:14 PM
Graham,
Did you mean 470R? My maths might be wrong:

V = 12
I = 25mA, .025A

R = 12/.025
= 480

if I got that wrong feel free to make fun :P

tango
06-01-2003, 07:16 PM
also, why 10k? I know a transistor needs .6v to operate, but I was just wondering what maths is used to calculate that.

Andy, I looked up optocouplers on the dse site, and they have 6 legs. what are they all for?

-=JM=-
06-01-2003, 08:54 PM
Surely you could just remove the case LED and put in your new one and it will just work ?

godfather
06-01-2003, 09:20 PM
Tango, transistors are current amplifiers, and whether the base drive was 1k or 10k will not matter, as either will drive the transistor to saturation.
A 10k resistor, even driven with 5v will allow 0.5mA to flow.
Most transistors of this type have an "amplification factor" of 200 or more, so 200 x 0.5 = up to 100 mA can flow in the collector.
As you are limiting it to 25 mA with the series resistor (47R), the transistor is fully saturated anyway, the best way to switch.
You do not size base current to meet the expected collector current only in a switching situation, as the power dissapation in the transistor will be excessive.

The 47R is based on 12v - 10.5v = 1.5 (as the junction voltage is 10.5 for the LED) NOT the full 12v.
47 ohms is close to what is needed. (1.5/.025=~60 ohms).

Graham is probably right that the blue LED is actually several LED's in series, hence the high Vf. Hence the need for a limiting resistor, as there is no internal resistor.

E&OE

Chilling_Silence
06-01-2003, 09:45 PM
> Surely you could just remove the case LED and put in
> your new one and it will just work ?

Yes, but some people have to have lights that dance and sing along to their HDD's workings :p

tango
06-01-2003, 09:55 PM
> Surely you could just remove the case LED and put in
> your new one and it will just work ?

I tried that, except the voltage coming out of it isn't enough to run the big LED :P

Chilling_Silence
06-01-2003, 10:07 PM
>
> I tried that, except the voltage coming out of it
> isn't enough to run the big LED :P

Ah, so you want your neighbours to know that your HDD is being defragged and working hard :p :D

*Chilling_Silence realizes he should not give up his day-job*

godfather
06-01-2003, 10:08 PM
The case LED will be a "normal" one with about Vf~2v.
The current limiting resistor will be ~ 470 or 1k depending on the voltage (5 or 12) for the case LED
The current through the blue LED (Vf~10.5v) would be ~ 1.5 ~ 3 mA, not enough for a glow-worm

Clueless
07-01-2003, 08:25 AM
I tried the link and got this:

SCR C106D / TYN404 400V 4Amps
T0202/T0126 Sensitive Gate
Cat No. Z4320
NZ\$2.70

Me thinks your cat number might be wrong....
What is this LED? Something designed for a car or something?

.Clueless

tango
07-01-2003, 08:36 AM
oh geeze..I'm really losing it today :P

it's actually Z4230, sorry :)

Clueless
07-01-2003, 09:00 AM
Cool LED!
I noted in th description:

"Blue LED that uses only 2 leads for simplicity of connection instead of the usual 12 found in LEDs of such sizes."

My suggestion is to just get a blue led with "the usual 12 found in LEDs of such sizes", which i would assume is 2 wires for each component LED, and then wire them up in 3 pairs so that the voltage is an adequate match...

.Clueless

Graham L
07-01-2003, 02:38 PM
Many of the component values in electronic designs are arbitrary. As godfather says, 10k or 1k for driving a transistor from logic will saturate (turn on fully) the transistor. I would normally use 10k or 4k7, depending on which bin was fullest.

I actually calculated (with envelope and pencil), the 47 ohm value (godfather left out the voltage across the saturated transistor ;-) --- so his value is high --- I got "about 50" so went to the nearest preferred value. From his value of "~60", we'd go to either 68 (or 47 ... depending on the stocks).

I don't see the point of using a optocoupler ... the circuit is all in the same box and using the same supplies.

I spent this morning working on a dead 600W inverter (60A from a 12V battery to make 230VAC). The clever manufacturers had polished off the type numbers on the transistors .X-( I drew the circuit a while ago for a friend who also had a non-working one. Mine half works. After desoldering the shorted transistors and looking at them a bit harder I have finally identified the type number ... they're \$50 each:_|. There are four of them. I think I will change from these transistors --- BUR50S (120V 70A) --- to power Mosfets at \$5 each. Amazing the difference in modern technology: the dissipation will be reduced by a factor of 50. But the change in design will need some proper calculations. I think I'll use my sliderule for technical things like that. :D

tango
07-01-2003, 07:51 PM
ok change of plan...I'm going to run the big 10.5v LED off a 12 volt supply now. I know I'll need to lower the voltage, but what about the current? how can I tell if it's less than 25mA?

And I'll get one of them superbrights and hook it up to the HDD light output. I think need another resistor for that but should I worry about current overload?

godfather
07-01-2003, 08:14 PM
> ok change of plan...I'm going to run the big 10.5v
> LED off a 12 volt supply now. I know I'll need to
> lower the voltage, but what about the current? how
> can I tell if it's less than 25mA?

If its more than 25 mA, you will tell as it will go for a few milliseconds only. Remember 25 mA is the MAXIMUM. Best to run slightly below.

12 volts - 10.5 volts = 1.5 volts
for 25 mA to flow with a drop of 1.5 volts across the resistor, its: 1.5 divided by .025 = 60 ohms (nearest should be higher to protect the LED, use two 33 ohm resistors in series) If they have the resistor range that has 62 ohms, then use it otherwise 2 x 33.

Deja Vu...I can remember doing this before?

> And I'll get one of them superbrights and hook it up
> to the HDD light output. I think need another
> resistor for that but should I worry about current

In place of the existing LED or ?

tango
07-01-2003, 08:30 PM
oh sorry, in place of the existing one thats there.

godfather
07-01-2003, 09:10 PM
Well, the standard HDD LED will probably have a lower Vf than a superbright.
Superbrights are around 3.6v Vf, 20 mA, a tad higher than the standard. Nowhere near the extreme of the blue one though.

You will be flat out trying to replace the resistor though, its bound to be a SMC (surface Mount Component) on the IDE electronics printed circuit board on the drive. Case one ditto, but on the motherboard.

Just use the resistor thats there, it means the new LED won't be quite as bright as it could be. Simply replace it with the superbright.

As with all these LED's take care to get them round the right way, first time.

-=JM=-
07-01-2003, 09:45 PM
I thought the wires which you plug onto the motherboard from the HDD LED in the front of the case had the resistance built into it.

godfather
07-01-2003, 09:51 PM
> I thought the wires which you plug onto the
> motherboard from the HDD LED in the front of the case
> had the resistance built into it.

You may be thinking of a resistor soldered into the wire and heat-shrinked up, but none of my recent ones have had this, just a connector on the mobo. This was common practice for 230v neon indicators which had a very high value soldered into the lead, for safety as these will explode if direct connected to the mains. (as will LED's) :D

The HDD LED on the actual HDD itself will have on-board resistor, and the case one will have a motherboard resistor in most situations based on my observations.

tango
07-01-2003, 09:57 PM
I'm not sure with anything to do with the actual HDD, I'm just thinking of using the connector that is on the mobo. I tested with my multimeter, and it was just below either 4 or 5 volts, I can't remember which. probably 4v. Hopefully it's about 3.6 :). The current coming out of the motherboard connector should be suitable for the superbright LED, right?

godfather
07-01-2003, 10:27 PM
Here we go again...

Was the case LED connected when you measured the voltage?
Is the case LED fed from 12v or 5v?

If the case LED was disconnected, but trying to operate, then the voltage you measured would be the full supply voltage irrespective of the resistor.

If the multimeter was the only load, the resistor would not make any real difference. A digital multimeter has an internal resistance of 10,000,000 ohms, which draws almost no current, which doesnt load the circuit. A cheap analogue meter will have about 20,000 ohms, and much the same result.

If the LED was the only load, and was operating, then the voltage you measure could be the Vf of the case LED.

Maths again.

If the case LED has a Vf of 2.2 volts and is fed from 12 volts, then your resistor in the circuit will be (12-2.2)/.02 = 490 ohms. Say 470 or anything to 1k (you dont HAVE to run the full 20 mA)

If the case LED is fed from 5v, then (5-2.2)/.02 = 140. Say 150 or so as a resistance value.

12v feed (12-3.6)/470 = .017 amps (17 mA) which would be fine

5v feed (5-3.6)/150 = 9mA, which would be a tad dull.

The last 2 equations assume a 470 or a 150 ohm resistor. The actual value could be higher, so your currents would be lower and output lower.

Ever heard of "suck it and see?"

Just remember you cannot run both LEDs in parallel. Series yes, but you will run out of volts doing that. Just try the superbright in place of the case LED. If its not satisfactory, measure the current flowing in the circuit.